Article pubs.acs.org/biochemistry
Alkylated Hydroxylamine Derivatives Eliminate Peripheral Retinylidene Schiff Bases but Cannot Enter the Retinal Binding Pocket of Light-Activated Rhodopsin Ronny Piechnick, Martin Heck, and Martha E. Sommer* Institut für Medizinische Physik und Biophysik (CC2), Charité-Universitätsmedizin Berlin, Charitéplatz 1, D-10117 Berlin, Germany ABSTRACT: Besides Lys-296 in the binding pocket of opsin, all-trans-retinal forms adducts with peripheral lysine residues and phospholipids, thereby mimicking the spectral and chemical properties of metarhodopsin species. These pseudophotoproducts composed of nonspecific retinylidene Schiff bases have long plagued the investigation of rhodopsin deactivation and identification of decay products. We discovered that, while hydroxylamine can enter the retinal binding pocket of light-activated rhodopsin, the modified hydroxylamine compounds o-methylhydroxylamine (mHA), o-ethylhydroxylamine (eHA), o-tertbutylhydroxylamine (t-bHA), and o-(carboxymethyl)hydroxylamine (cmHA) are excluded. However, the alkylated hydroxylamines react quickly and efficiently with exposed retinylidene Schiff bases to form their respective retinal oximes. We further investigated how t-bHA affects light-activated rhodopsin and its interaction with binding partners. We found that both metarhodopsin II (Meta II) and Meta III are resistant to t-bHA, and neither arrestin nor transducin binding is affected by t-bHA. This discovery suggests that the hypothetical solvent channel that opens in light-activated rhodopsin is extremely stringent with regard to size and/or polarity. We believe that alkylated hydroxylamines will prove to be extremely useful reagents for the investigation of rhodopsin activation and decay mechanisms. Furthermore, the use of alkylated hydroxylamines should not be limited to in vitro studies and could help elucidate visual signal transduction mechanisms in the living cells of the retina.
R
hodopsin (λmax = 500 nm) consists of the chromophore 11-cis-retinal covalently linked via a protonated Schiff base to Lys-296 of the G-protein-coupled receptor opsin.1,2 The retinylidene Schiff base in rhodopsin is extraordinary in its chemical stability,3 which is bestowed by the protein structure.4 Light catalyzes the isomerization of 11-cis-retinylidene to all-trans-retinylidene, which induces conformational changes in the protein structure and the evolution of a series of metarhodopsin species.4 Within milliseconds, metarhodopsin II (Meta II; λmax = 380 nm) evolves in equilibrium with its precursor Meta I (λmax = 480 nm). Notably, the retinylidene Schiff base in Meta II is deprotonated, and Meta II can couple to and activate the G-protein transducin. Meta II is not stable and decays within minutes as the retinylidene Schiff base is hydrolyzed and all-trans-retinal is released.5 In addition, lightactivated metarhodopsin can decay via Meta III (λmax = 470 nm). Meta III results from anti → syn isomerization of the retinylidene Schiff base, which can occur spontaneously in Meta I6,7 or can be induced in Meta II by blue light.8 Meta III is relatively long-lived (minutes to hours) but eventually decays to opsin and free all-trans-retinal (for a review, see ref 9). Historically, the investigation of the rhodopsin activation and decay cycle has been complicated by the chemical reactivity of retinal. Retinal reacts with amines to form Schiff bases, particularly those found in phosphatidylethanolamine, 10,11 phosphatidylserine,12 and cytoplasmically exposed lysine residues of opsin.13,14 On the basis of the reported pKa of 7.3 for model retinylidene Schiff bases in solution, 3 approximately half of these peripheral Schiff bases would be © 2011 American Chemical Society
protonated at physiological pH. Because the absorbance of a protonated Schiff base (λmax = 440 nm)15 is similar to those of many metarhodopsin species, these nonspecific Schiff bases have often been mistaken for genuine photoproducts (and vice versa). Thus, we call these nonspecific retinylidene Schiff bases pseudophotoproducts. For instance, the existence of Meta III was long debated, with many investigators believing it was simply a pseudophotoproduct (for a review, see ref 16). The strong nucleophile hydroxylamine (HA) (Figure 1A) has long been used to rapidly cleave the retinylidene Schiff base in Meta II and convert retinal to retinal oxime. In addition, HA has been useful in probing the relative stability of rhodopsin mutants17,18 and in elucidating how the bulk solvent enters Meta II to hydrolyze the retinylidene Schiff base.19 The alkylated hydroxylamine derivatives o-methylhydroxylamine (mHA) and o-ethylhydroxylamine (eHA) (Figure 1A) have been used for some time to convert retinal to alkylated retinal oxime for high-performance liquid chromatography (HPLC) analysis.20−22 We report here that these alkylated hydroxylamines, in addition to o-tert-butylhydroxylamine (t-bHA) (Figure 1A), can be used to preferentially cleave peripheral Schiff bases while not affecting Meta II. Furthermore, we show that t-bHA does not affect Meta III or transducin and arrestin binding to Meta II. This biochemical tool allows for the straightforward discrimination of pseudophotoproducts and Received: May 3, 2011 Revised: June 29, 2011 Published: July 18, 2011 7168
dx.doi.org/10.1021/bi200675y | Biochemistry 2011, 50, 7168−7176
Biochemistry
Article
ROS isolated from retinas originating from the United States were used to prepare phosphorylated rhodopsin using the native rhodopsin kinase, following an earlier protocol25 that was later modified.26 After regeneration of phosphorylated ROS membranes with 11-cis-retinal, free retinal was removed with extensive BSA washes as described previously. 26 ROS membranes that were not phosphorylated were prepared from locally obtained retinas, and soluble and membraneassociated proteins were removed by washing the membranes several times with low-ionic strength buffer [5 mM Pipes (pH 7) and 1 mM EDTA] as previously described.27,28 Opsin was prepared by bleaching these ROS in the presence of hydroxylamine, followed by repeated washes as previously described.29 Transducin was prepared from locally obtained retinas following an established protocol. 30 Recombinant mutant arrestin I72C, which lacked native cysteine and tryptophan residues, was expressed, purified, and labeled with the IANBD fluorophore as recently described.26 Note that labeled arrestin I72C is termed I72NBD. Preparation of Meta III. Meta III was formed in washed ROS membranes according to a published protocol,28,31 and Meta III yields were typically ∼20% of the total receptor yield. Membranes were subsequently solubilized in 0.1% n-dodecyl β-D-maltopyranoside (receptor concentration of approximately 3 μM) for 5 min at room temperature, followed by high-speed centrifugation (100000g for 10 min). Detergent-solubilized Meta III was stored on ice, and under these conditions, decay was slow enough that the sample could be used for experiments over several hours. Absorbance Spectroscopy. UV−visible absorbance spectra of samples were recorded in 1 cm path length quartz microcuvettes (90 μL) using a Varian Cary 50 spectrometer (2 nm resolution, 11 nm/s scan rate). For samples of washed ROS membranes, opsin membranes with light scattering properties that closely matched those of the ROS were used as a reference. Absorbance spectra were subsequently modified by subtracting an arbitrarily calculated curve that approximated the sloping baseline. Because the calculated baseline did not perfectly match the sample, the apparent absorbance maxima were wavelength-shifted. However, these shifts in absorbance do not affect the interpretation of our results. For samples of detergent-solubilized Meta III, sample buffer was used as the reference, and absorbance spectra were automatically recorded every 30 s. Difference spectra were calculated by manually subtracting spectra in SigmaPlot 10.0 (SysStat Software). Fluorescence Spectroscopy. Fluorescence measurements were taken on a SPEX Fluorolog (1680) instrument in the front-face mode. Membrane suspensions were stirred in a 1 cm path length cuvette (750 μL) equipped with a stir bar, and the temperature of the cuvette holder was controlled with a circulating water bath. Excitation slits were generally set between 0.1 to 0.2 nm, while emission slits were set at 4 nm. Meta II decay was monitored as an increase in opsin tryptophan fluorescence (λex = 295 nm; λem = 330 nm).31−33 For measuring steady-state emission spectra (500−660 nm) of I72NBD, we placed samples in a 0.3 cm path length microcuvette (80 μL) and excited them at 360 nm (2 nm step, 0.5 s integration per point). For time-based fluorescence acquisition, stirred samples (750 μL) were excited at 500 nm, and emission was measured at 550 nm. Because this excitation wavelength overlaps with the absorption of rhodopsin, excitation slits were reduced to 0.05 nm. Samples were exposed
Figure 1. Influence of HA and HA derivatives on the absorption properties of rhodopsin and Meta II. (A) Molecular structures of water (control), hydroxylamine (HA), o-methylhydroxylamine (mHA), o-ethylhydroxylamine (eHA), o-tert-butylhydroxylamine (t-bHA), and o-(carboxymethyl)hydroxylamine (cmHA). Space-filling models were rendered in Jmol (http://www.jmol.org/) using published van der Waals radii.61 Atoms are colored as follows: gray for carbon, red for oxygen, blue for nitrogen, and white for hydrogen. (B) Absorbance spectra of ROS membranes containing 1 μM rhodopsin were measured in the absence (black) or presence of 25 mM HA (blue), mHA (orange), eHA (green), t-bHA (red), or cmHA (magenta). Illumination in the absence of hydroxylamine resulted in the characteristic 380 nm absorbance peak of Meta II. The presence of mHA, eHA, t-bHA, or cmHA had no effect on Meta II formation. In contrast, illumination in the presence of HA resulted immediately in a 360 nm peak indicating retinal oxime. Spectra were recorded as described in Experimental Procedures, and rhodopsin was illuminated in the presence of 300 μM Gt α peptide to prevent Meta I formation.
metarhodopsin species. Furthermore, HA and t-bHA together represent a tool kit that can be used to easily discriminate among Meta II, Meta III, and opsin. We discuss how the application of this tool kit to living cells might help shed light on the molecular mechanisms of vision.
■
EXPERIMENTAL PROCEDURES
Materials. Bovine retinas were obtained either from a local slaughterhouse or from W. L. Lawson Co. (Omaha, NE). 11-cis-Retinal was created in house using commercially available all-trans-retinal and purified by HPLC.23 The high-affinity analogue peptide derived from the α-subunit of transducin (G t α peptide), VLEDLKSCGLF, was synthesized by P. Henklein (Institut für Biochemie, Charité). Chromatography supplies for arrestin purification were purchased from GE Healthcare, and the fluorescent probe IANBD [N,N′-dimethylN-(iodoacetyl)-N′-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)ethylenediamine] was from Invitrogen. All other reagents were from Sigma or Fluka. Membrane and Protein Preparations. Rod outer segments (ROS) were prepared from frozen retinas using the previously described discontinuous sucrose gradient method. 24 7169
dx.doi.org/10.1021/bi200675y | Biochemistry 2011, 50, 7168−7176
Biochemistry
Article
to a bright light source (495 nm long-pass filter. For light scattering measurements, transducin and arrestin binding to the membrane was initiated with a flash of light (
